Skip to content
SPECIFICTION
Get listed

MinerU

opendatalab.github.io

From $4.99/mo Free plan Open source WebWindowsmacOSLinuxAPISelf-hosted
Visit opendatalab.github.io
Spec score 9.1 #3 of 47 in PDF Converter Software

What it is

MinerU 4.0 combines document parsing, a local document library, and tools for conversion, application integration, and agent reading. It accepts PDFs, images, office documents, ebooks, web formats, and other listed inputs, and can produce Markdown, HTML, LaTeX, DOCX, EPUB, PDF, structured content, and content lists. Four parsing tiers—Flash, Basic, Standard, and Advanced—serve different quality and compute needs. The library can discover files, cache parsing results, search content, continue by page or block, and preserve citation locators. Developers can use a Python SDK, V1 HTTP API, command-line tools, or a Gradio-based WebUI. MinerU supports the MCP protocol and agent frameworks, with exports to Dify and Notion. Documents are not automatically uploaded to the official service; remote parsing requires explicit configuration. Local WebUI and API server deployments are documented for Linux, macOS, and Windows. The open-source software is free under its project license; commercial-use limits apply above 100 million monthly active users or USD 20 million monthly revenue.

Spec sheet

What it does
MinerU 4.0 combines document parsing, a local document library and service tools for conversion, application integration and agent reading.Read 15 Oct 2026 · opendatalab.github.io
Purpose
MinerU 4.0 combines document parsing, a local document library, and service tools for conversion, application integration, and agent reading.Read 5 Oct 2026 · opendatalab.github.io
Support
The FAQ directs users with unresolved problems to the MinerU Discord or WeChat community.Read 5 Oct 2026 · opendatalab.github.io
License
MinerU uses a custom MinerU Open Source License based on Apache 2.0, with attribution requirements and a separate commercial-license trigger above 100 million monthly active users or 20 million USD monthly revenue.Read 15 Oct 2026 · github.com
Agent integrations
The official site says MinerU supports the MCP protocol and agent frameworks, and offers export to Dify and Notion.Read 5 Oct 2026 · mineru.net
API network limitation
The hosted create-task API documentation warns that URLs hosted on GitHub, AWS and other overseas services may time out.Read 15 Oct 2026 · mineru.net
API rate limits
The hosted service documents 300 submit requests per minute, 1,000 result queries per minute and up to 10,000 files per day, subject to dynamic adjustment.Read 15 Oct 2026 · mineru.net
Changelog license change
Release 3.1.0 dated April 18, 2026 changed the project from AGPLv3 to the custom Apache-2.0-based MinerU Open Source License.Read 15 Oct 2026 · opendatalab.github.io
Cloud lightweight API
The Agent Lightweight API requires no login but is IP-rate-limited, handles one file up to 10 MB and 20 pages, and returns Markdown only.Read 15 Oct 2026 · mineru.net
Cloud precision API
The Precision Extract API requires a token, accepts files up to 200 MB and 600 pages, batches up to 200 files, and returns ZIP results with Markdown and JSON plus optional DOCX, HTML or LaTeX.Read 15 Oct 2026 · mineru.net
Company headquarters and size
Crunchbase lists OpenDataLab as a private for-profit organization headquartered in Shanghai with 51–100 employees.Read 15 Oct 2026 · crunchbase.com
Company identity
OpenDataLab's official site identifies it as the Data Platform Center of Shanghai Artificial Intelligence Laboratory and the maker of MinerU.Read 15 Oct 2026 · opendatalab.github.io
Developer tooling
Official tooling includes a Python SDK, V1 HTTP API, stateless batch conversion, multi-service Router and Gradio WebUI.Read 15 Oct 2026 · opendatalab.github.io
Docker support
The documented Docker workflow targets Linux NVIDIA GPUs and Windows WSL2; macOS Docker has no MPS acceleration and non-NVIDIA deployment is pending.Read 15 Oct 2026 · opendatalab.github.io
Document library
Its library discovers and caches files, supports content search and continuation by page or block, and preserves stable citation locators.Read 15 Oct 2026 · opendatalab.github.io
Export formats
The document model supports Markdown, HTML, LaTeX, DOCX, EPUB, PDF, Structured Content and Content List V1/V2 outputs.Read 15 Oct 2026 · opendatalab.github.io
Hardware requirements
The docs recommend at least 16 GB system memory for higher-throughput local deployments and at least 16 GB unified memory for Apple Silicon.Read 5 Oct 2026 · opendatalab.github.io
Hosted service
The official site links to online use and describes an official cloud service, but the opened pages did not state its pricing.Read 5 Oct 2026 · mineru.net
Input formats
Supported inputs include PDF, images, DOC/DOCX, PPT/PPTX, XLS/XLSX, RTF, OpenDocument, EPUB, OFD, HTML/MHTML and CSV/TSV.Read 15 Oct 2026 · opendatalab.github.io
Installation
The base package is installed with mineru>=4.0,<5 and includes the WebUI, ONNX runtime and llama.cpp VLM.Read 15 Oct 2026 · opendatalab.github.io
Institution founding
LinkedIn lists Shanghai Artificial Intelligence Laboratory as founded in 2020, headquartered in Shanghai, with 201–500 employees.Read 15 Oct 2026 · linkedin.com
Integration
MinerU provides a Python SDK, V1 HTTP API, command-line tools, and a Gradio-based WebUI.Read 5 Oct 2026 · opendatalab.github.io
License terms
The project license permits commercial use without a separate license unless monthly active users exceed 100 million or monthly revenue exceeds USD 20 million, and requires attribution for third-party online services based on MinerU.Read 5 Oct 2026 · github.com
Limit: active defect areas
The repository bug list includes reports involving Apple MPS OCR falling back to CPU, formula and punctuation errors, figure provenance, table extraction and Arabic recognition/order.Read 15 Oct 2026 · github.com
Limit: backend quality gap
DocuOCR reports the CPU pipeline backend at 86.47 on the same benchmark, about 8.92 points below the VLM result.Read 15 Oct 2026 · docuocr.com
Limit: CPU timeout
A closed GitHub issue reports that pure CPU mode could not finish PDFs over 200 pages before timing out.Read 15 Oct 2026 · github.com
Limit: hosted throughput
DocuOCR says hosted accounts receive 1,000 pages per day at highest priority and lower priority thereafter, with 200 MB and 200-page per-file limits.Read 15 Oct 2026 · docuocr.com
Limit: slow OCR preprocessing
A GitHub discussion records a report of more than five minutes of preprocessing for a 120-page OCR PDF, attributed to CPU-bound PDF-to-image conversion.Read 15 Oct 2026 · github.com
Local privacy behavior
Local documents are not automatically uploaded to the official service; remote parsing requires explicit configuration.Read 15 Oct 2026 · opendatalab.github.io
Local processing
The docs say documents are not automatically uploaded to the official service; remote parsing requires explicit configuration.Read 5 Oct 2026 · opendatalab.github.io
Model runtimes
Small models can use ONNX or Torch, while VLM inference can use llama.cpp, vLLM or LMDeploy.Read 15 Oct 2026 · opendatalab.github.io
On-premise security
The hosted product advertises fully on-premise deployment so models and data stay inside the customer's network, with a compliance-ready framework and dedicated support.Read 15 Oct 2026 · mineru.io
Operating systems
The current quick start gives installation instructions for Linux, macOS, and Windows environments.Read 5 Oct 2026 · opendatalab.github.io
Optional packages
The torch extra adds Torch, Torchvision, Transformers and Accelerate; the full extra adds vLLM on Linux or LMDeploy on Windows.Read 15 Oct 2026 · opendatalab.github.io
Outputs
Its shared document model supports Markdown, HTML, LaTeX, DOCX, EPUB, PDF, Structured Content, and Content List V1/V2 outputs.Read 5 Oct 2026 · opendatalab.github.io
Parsing tiers
The software provides Flash, Basic, Standard and Advanced parsing tiers for progressively more demanding speed, OCR and layout-quality needs.Read 15 Oct 2026 · opendatalab.github.io
Praise: benchmark accuracy
DocuOCR reports MinerU2.5-Pro scoring 95.75 on OmniDocBench v1.7, ranked second among 32 systems.Read 15 Oct 2026 · docuocr.com
Praise: open source and community
AIToolCrux praises the free availability and active community, while its editorial advantages list highlights documentation and RAG support.Read 15 Oct 2026 · aitoolcrux.com
Python requirement
MinerU 4.0 supports Python versions 3.10 through 3.14, with Python 3.12 recommended for a new environment.Read 15 Oct 2026 · opendatalab.github.io
Rating on AIToolCrux
8.0 out of 10 from one editorial evaluation; the page gives a B grade and says functionality scored 9.2/10.Read 15 Oct 2026 · aitoolcrux.com
Security policy
The repository security policy supports vulnerability reports through GitHub Security Advisories and asks users to avoid public disclosure before fixes.Read 15 Oct 2026 · github.com
Self-hosted service
A local WebUI runs on port 7860 and the local API server on port 8000 with OpenAPI documentation.Read 15 Oct 2026 · opendatalab.github.io
Self-hosting
The documentation describes running a local WebUI and V1 API server, and connecting the SDK to a self-hosted API.Read 5 Oct 2026 · opendatalab.github.io

Plans and prices

Free Free $0/ month 5 free pages every day · files up to 10 MB · images up to 5 MB · PDFs up to 5 pages · web access only, no API Read 15 Oct 2026 · mineru.io
MinerU open-source software Free Free to use under the project license Commercial license required above 100 million monthly active users or USD 20 million monthly revenue · online services must attribute MinerU Read 5 Oct 2026 · github.com
Pro · Monthly $9.90 a month USD · one-time payment 1,000 pages every month · files up to 200 MB · images up to 50 MB · PDFs up to 500 pages · API included Read 15 Oct 2026 · mineru.io
Pro · Annual $59.90 a year USD · one-time payment 1,000 pages every month · files up to 200 MB · images up to 50 MB · PDFs up to 500 pages · API included Read 15 Oct 2026 · mineru.io
On-premise Priced on request Talk to us about deployment models and data run inside your network · dedicated support Read 15 Oct 2026 · mineru.io

Every MinerU plan and price →

Strong points and limits

Strong points

  • Accepts a broad range of document formats.
  • Offers four parsing tiers.
  • Provides SDK, HTTP API, CLI, and WebUI.
  • Documents are not uploaded automatically.

Limits

  • Higher-throughput local deployments recommend at least 16 GB system memory.
  • Apple Silicon deployments recommend at least 16 GB unified memory.
  • Commercial-use limits apply above stated thresholds.

Who it suits

It suits people building document conversion or parsing workflows, including developers and users integrating with agents. Teams can run a local WebUI or API server on Linux, macOS, or Windows.

Verdict

MinerU brings document parsing, library functions, and integration tools into one open-source project. Its local processing options and tiered parsing are useful distinctions; commercial users should note the license thresholds.

Where Specifiction places it

Alternatives to MinerUAll alternatives →